Time Nick Message 01:39 RealBadAngel i added collision boxes to nodedef 01:39 RealBadAngel i can merge that into meshes pull or make a separate commit with it after meshes got merged 01:41 VanessaE separate pull I think 02:14 paramat hmmmm, i have the mgv7 biome API working in mgv5 https://github.com/paramat/minetest/commits/mgv5 perhaps sometime you could review my commits? https://cdn.mediacru.sh/EHEmPPVfI0FS.png 02:31 paramat big freaking dig mapgen in mgv5 https://cdn.mediacru.sh/YMssAZefdaBQ.png 03:22 OldCoder http://minetest.org/141017-patch.txt 03:22 OldCoder Zeno`, ^ as promised. Other devs, please consider commenting on negative sequence numbers. 03:22 OldCoder Those do not seem normal 03:39 hmmmm hey paramat, I will do that right away but not tonight 03:39 hmmmm sorry 03:40 hmmmm i've been gone all week on a business trip and had literally no time for anything 03:41 hmmmm i'm gonna need to put some of my minetest todos on hold and get biomes fixed because right now they're a bunch of shit 03:41 hmmmm and the primary reason why there aren't any biomes defined by default 04:04 paramat hmmmm, no problem 04:07 paramat yes some basic biomes for mgv7 and eventually mgv5 would be good, i intend to get needles, pine treetrunk and pinewood nodes added to default so we can have snowy biomes too 06:37 RealBadAngel hmmm, opinions on https://github.com/minetest/minetest/pull/1738 ? 06:39 RealBadAngel hmmmm, ^ 08:49 RealBadAngel no other devs around? 09:00 RealBadAngel 1738 is complete, i need to get that merged for another stuff - i do have custom collision boxes pull ready too, but it depends on previous one 09:01 RealBadAngel i quit my last job and will start another next month, so i have a bit spare time 09:02 RealBadAngel i can code some stuff now but need it merged 10:15 RealBadAngel kahrl, here? 10:42 RealBadAngel celeron55, ? 10:43 RealBadAngel sfan5 ? 10:43 RealBadAngel ShadowNinja, ? 10:43 Zeno` RealBadAngel ? 10:43 RealBadAngel youre not a dev yet ;) 10:44 Zeno` Didn't say I was. I just wanted to say your name 10:44 RealBadAngel lol 10:44 RealBadAngel shit i dont like sitting unproductive 10:44 Zeno` re-organising the_game() is 70% complete 10:45 RealBadAngel its not green anymore? ;) 10:45 Zeno` wield hand seems to get stuck waving and I haven't added profiler graph back yet 10:45 Zeno` no it's not green now :) 10:45 Zeno` I can connect and play (sort of) now 10:45 PenguinDad Zeno`: in how many moons will it be finished? :D 10:45 Zeno` oh I need to add status text back as well 10:46 Zeno` PenguinDad, probably tomorrow 10:46 PenguinDad nice 10:47 Zeno` maybe the day after.. it's still a mess 10:47 Zeno` https://github.com/Zeno-/minetest/blob/refactor_the_game/src/game.cpp#L2072 10:48 Zeno` well, 2129 is basically the loop 10:48 Zeno` I need to move that other stuff somewhere else at some point 10:49 Zeno` err ignore line 3610 :-o 10:49 Zeno` that's just temporary ROFL 10:53 sol_invictus is it possible to determine if an entity (player) is in a free fall? 11:04 iqualfragile sol_invictus: yes: if True: 11:09 sol_invictus iqualfragile: very helpful 11:10 iqualfragile sol_invictus: what are you trying to archive? 11:10 iqualfragile don't you want to check the velocity? 11:11 sol_invictus that's that came to my mind first 11:12 sol_invictus just not sure if it's the simplest way 13:43 RealBadAngel PilzAdam, are you ok with mergeing meshnodes pull? 13:43 PilzAdam haven't looked at the code 13:44 RealBadAngel if you could then 13:44 PilzAdam but I generally agree to the idea that nodes can have meshes 13:44 PilzAdam I don't have time for this now, sorry :-/ 13:44 RealBadAngel kahrl already reviewed the code 13:45 RealBadAngel he had some suggestions and i already changed the code 13:46 RealBadAngel PilzAdam, point is i have days off, i do have pulls waiting in line already, and cant code anything new 13:46 RealBadAngel im blocked for 2 days already 13:47 RealBadAngel since 2 days im sitting here and just waiting 13:57 RealBadAngel hi nore 13:57 nore hi 13:57 RealBadAngel do you have some time to review meshnodes? 14:00 nore hm.. sorry, not now 14:03 RealBadAngel damn, anybody then? 14:03 RealBadAngel whats wrong with you guys??? 14:03 proller 8) 14:05 Krock RealBadAngel, I like it. 14:09 Zeno` Does anyone have time to review #1739 14:09 ShadowBot https://github.com/minetest/minetest/issues/1739 -- Move pause menu buttons upwards to accommodate for new configure keys button by Zeno- 14:17 sfan5 Zeno`: looks good 14:17 RealBadAngel sfan5, what about you? 14:17 Zeno` cool, I'll leave it then 14:17 sfan5 RealBadAngel: ? 14:17 RealBadAngel https://github.com/minetest/minetest/pull/1738 14:18 RealBadAngel Zeno`, you have already 2 votes. will merge your pull in a moment 14:19 sfan5 RealBadAngel: https://github.com/minetest/minetest/pull/1738/files#r19053887 14:19 Zeno` oh ok. Thanks 14:20 RealBadAngel sfan5, what wrong with style here? 14:21 RealBadAngel not to mention its a copy of existing code 14:22 sfan5 RealBadAngel: "In case your function parameters don't fit [...]. Indention for follow up lines is EXACTLY two tabs." 14:22 sfan5 for(std::vector::iterator 14:22 sfan5 + i = boxes.begin(); 14:22 sfan5 .. 14:22 sfan5 is 4 tabs, not 2 14:22 sfan5 and on line 542 it is indented without a reason 14:25 RealBadAngel sfan5, indeed, fixed it 14:25 RealBadAngel anything more? 14:25 sfan5 *looks* 14:26 sfan5 fixed it, where? 14:26 RealBadAngel local, shall i push it? 14:26 RealBadAngel i was just waiting for anything else 14:26 sfan5 yes 14:26 sfan5 well 14:26 sfan5 I can't tell if you missed anything without seeing the changes 14:27 RealBadAngel ok, plz hold on a sec 14:28 RealBadAngel pushed 14:30 sfan5 looks good 14:31 RealBadAngel are you ok with commit then? 14:31 sfan5 sure 14:32 RealBadAngel good. and thx 14:32 RealBadAngel in a 15 minutes im going back with next one 15:24 iqualfragile yay, meshnode drawtypes! 15:24 iqualfragile thanks RealBadAngel! 16:03 kahrl since the in game key change menu was merged (or even before?), the mouse cursor is no longer on the disconnect button when you open the pause menu 16:03 kahrl this should be fixed 16:12 kahrl actually that might have happened when the pause menu was changed to be a formspec 16:14 kahrl ok, I guess if nobody complained since then, nobody relied on that feature 16:38 Calinou “Comments should follow English grammar rules, this means starting with a capital letter, using commas and apostrophes where appropriate, and ending with a period.” → great 16:38 Calinou exactly what I'm doing :P 16:38 Calinou British English or American English? 16:42 PenguinDad Calinou: Most likely American 16:47 RealBadAngel https://github.com/minetest/minetest/pull/1746 16:47 RealBadAngel ^^ this adds custom collision boxes for nodes 16:47 RealBadAngel defined the same way as nodeboxes 16:49 rubenwardy Comments? Only set the camera update keymap when using a debug build - https://github.com/minetest/minetest/pull/1744 16:54 gravgun Node meshes are merged, yay :) @sfan5: is the custom drawtype VM still a good idea? 17:16 sfan5 gravgun: possibly 17:16 gravgun Generally the VM works 17:16 gravgun Haven't secured some over/undlerflows though 17:17 gravgun Right now I'm working on ProgramState saving/loading, Program save/load is already done 17:17 gravgun Well IDK if we'll need ProgramState (de)serialization 17:17 Calinou ^^ this adds custom collision boxes for nodes 17:17 Calinou thanks thanks thanks thanks thanks 17:17 Calinou now we can have proper non-jumpable fences and such :P 17:17 Calinou without small nodebox hacks 17:18 Calinou propety → property though 17:18 RealBadAngel fixed already 17:18 RealBadAngel just edited that ;) 17:18 Calinou problem with your nodebox → meshnode conversion, RealBadAngel: if you spin nodes using screwdriver, the lighting on them is spun too… 17:19 rubenwardy Calinou: https://cdn.mediacru.sh/2A-0DjVpbj8B.png 17:19 rubenwardy Is that better? 17:19 rubenwardy The text doesn't stand out with yellow 17:20 RealBadAngel rubenwardy its a static text? 17:20 rubenwardy Huh 17:20 rubenwardy It uses fgettext, so can be tranlated 17:20 rubenwardy Oh, it is label[] 17:20 RealBadAngel i mean propably using a tooltitp 17:20 RealBadAngel that would be better imho 17:21 RealBadAngel Calinou, i wont fix that propably 17:21 RealBadAngel we are a few steps to get real lighting 17:22 Calinou https://cdn.mediacru.sh/m6Uqkkr2BrS6.png 17:22 Calinou RealBadLighting :p 17:22 RealBadAngel 500 Internal Server Error 17:22 Calinou rubenwardy, yes, good 17:22 RealBadAngel i coded that? ;) 17:22 Calinou way less dramatic 17:23 Calinou I understand it doesn't stand out with yellow 17:23 Calinou RealBadAngel, that bug is almost a blocker :/ 17:23 RealBadAngel not rly 17:23 RealBadAngel i fixed that issue already 17:24 RealBadAngel now have to do that again, thx to caching 17:24 RealBadAngel not a big problem tho 17:24 RealBadAngel its a few liner 17:24 RealBadAngel and code to do that is already in core 17:25 Calinou I get a good 130 - 220 FPS with view range 130 currently 17:25 Calinou (placed a lot of slabs) 17:25 rubenwardy 22 FPS ? 17:25 rubenwardy * 220 FPS? 17:25 RealBadAngel try to place a lot of mushrooms 17:26 RealBadAngel anyway theres no difference between meshes and nodeboxes now 17:26 RealBadAngel both will need more time when more complicated 17:26 proller Calinou proof screenshot with f5 f5 info 17:27 RealBadAngel but stairs for example on my box are more than 10x faster 17:28 RealBadAngel now r2d2 model is as quite fast as stairs were before 17:28 PenguinDad Dreambuilder is much faster with that nodebox -> mesh conversion 17:28 rubenwardy Well done, RBA 17:28 rubenwardy I have gone from 35 - 55 to 50 - 60 17:28 rubenwardy in FPS. 17:29 rubenwardy Or it might be chance 17:29 RealBadAngel theres still lots to do 17:50 Calinou proller, https://cdn.mediacru.sh/JJY4I6S4LFiA.png 17:50 Calinou less high than I thought in forest 18:03 Sokomine hm. why does my mt version number stick to 0.4.9, although git seems to think it's up to date? (apart from desired local modifications). also, git status says "Your branch is ahead of 'origin/master' by 465 commits." while it should be equal to master (apart from the tiny local modifications) 18:03 rubenwardy git fetch ? 19:18 kilbith fbfg 19:37 Calinou can we use texturable sun and moon in game textures now (outside of texture packs) 19:38 proller Calinou, but on your screen not real 130 range 19:38 Calinou walked around to load terrain 20:00 RealBadAngel Calinou, the issue with lighting is trivial 20:01 RealBadAngel http://irrlicht.sourceforge.net/docu/classirr_1_1scene_1_1_i_mesh_manipulator.html#a8dba16b7eedeac07761c9c1247944429 20:01 RealBadAngel that has to be added after rotation 20:08 Sokomine RealBadAngel: ah, that would be good. i just checked with a small hut, and the lighting problem became obvious. though it did look well on its own 20:08 RealBadAngel its shading the faces problem 20:08 Sokomine it's very nice that we have those meshes now :-) hope the roofs will soon be changed 20:09 RealBadAngel meshes are copied, rotated but i forgot to recalculate normals 20:09 Sokomine well, i rotated a wooden planks slab upside-down. the wood became darker. which didn't look bad in that case 20:09 RealBadAngel and shading is basing on it 20:10 Sokomine ah, ok. then hopefully other devs can approve of your patch so that it may get added 20:11 RealBadAngel this is bugfix, i will merge it as soon as previous one gets it 20:12 RealBadAngel https://github.com/minetest/minetest/pull/1746/ 20:12 RealBadAngel this one waits in line 20:13 RealBadAngel btw, nodes are using node_box, collision_box, selection_box 20:14 RealBadAngel why objects use selectionbox ? 20:20 Calinou I'd like ability to use different collision_box, selection_box for entities 20:20 Calinou currently it's collisionbox, not collision_box 20:21 Sokomine hm, yes, collision box does not have to be as detailed as selection box. someone said it'd be a problem if there are too many/detailed collusion boxes 20:33 RealBadAngel nodeboxes do copy their boxes for selection boxes 20:33 Calinou your PR lets nodeboxes have separate collision and selection boxes, which is a good thing 20:34 Calinou (too many collision boxes cause glitches) 20:34 RealBadAngel yes 20:34 Calinou there should be as few as possible selection and collision boxes 20:34 RealBadAngel now modder can simplyfy it in case of complex ones 20:34 Calinou :( I've been looking in Git history, I see no mention of SmoothTranslator being added 20:34 Calinou (the process) 20:39 OldCoder The RE-SEND RELIABLE lockup strikes *again* 20:40 * OldCoder is very appreciative of any help for this issue which is arguably the most serious bug MT has 20:40 Calinou but no one of us had it ;) 20:40 OldCoder Others outside have had it 20:40 OldCoder And the log suggests a genuinely serious issue 20:40 OldCoder http://minetest.org/lockup.txt 20:40 OldCoder Tell me this is normal ^ 20:41 OldCoder I have now tried multiple patches and updating git and random poking at the code 20:41 OldCoder I have even tried ripping out Sapier commit as proller suggested 20:41 OldCoder But could not get the latter version to compile 20:41 OldCoder I have dozens of players eager for the world and no world to give them 20:42 OldCoder This is the most significant issue I have seen with the code in 2.5 years here 20:47 jin_xi maybe its an issue with the world? 20:47 OldCoder If so, it is an issue with the world that is triggering a core bug 20:47 OldCoder I have millions of resends 20:47 OldCoder Most who review the logs agree that this is very unusual 20:47 jin_xi i know 20:48 OldCoder I think that if something happens at exactly the wrong time, something is corrupted 20:48 OldCoder But it must be too fast or too slow just right; therefore, it is not reproducible 20:48 OldCoder But what *can* cause millions of resends? 20:48 OldCoder Surely the large number is a clue 20:48 * OldCoder has literally millions and millions of resends 20:48 jin_xi no need to resend that info 20:49 OldCoder Hm? 20:49 OldCoder Resend which info? 20:49 OldCoder Looks like 925,000 resends so far for today 20:49 jin_xi you stated the issue many times 20:50 OldCoder D: 20:50 OldCoder thank you for the informative remark 20:50 OldCoder I feel that periodic requests for help are reasonable jin_xi 20:50 OldCoder I have made my own contributions to the project for 2.5 years, small though they may be 20:50 jin_xi Hey dont be offended 20:51 jin_xi it was just a bad joke 20:51 jin_xi i apologize 20:51 OldCoder If you are offended that a contributor seeks to be helpful to the kids, this is your privilige 20:51 OldCoder Accepted 20:51 OldCoder But understand, I am behind in real life work that I need badly 20:51 OldCoder I had a flood of actual kids to this world and do not wish to disappoint them 20:51 jin_xi but my suggestion would be at this point to go look at the world in question and maybe methodically make it smaller/simpler to see if something changes 20:52 OldCoder jin_xi, this is RedCrab. Are you familiar with the world? 20:52 jin_xi oh. i see 20:52 OldCoder yes 20:52 OldCoder I suppose I will go poke at the code. But one position is firm: 20:52 OldCoder It is foolish 20:52 OldCoder for the project to have code that is so important, the connection code 20:52 OldCoder That only one person understands 20:53 OldCoder A project that does this is not viable in the long term. Believe me when I say that I have experience with this point. 20:53 OldCoder 20:53 Miner_48er Who develops this area of the core? 20:53 OldCoder Sapier, apparently 20:53 OldCoder I have looked and looked and not seen him 20:54 OldCoder For 1.5 weeks 20:54 Amaz Does this happen only on the SilverCrab world? 20:54 OldCoder I think it happens to others, possibly including you, but much more rarely 20:54 OldCoder Many factors may be involved including size 20:54 OldCoder This is a very large world and I still have 30GB to add 20:55 Miner_48er what's the current size? 20:55 OldCoder Hang on 20:55 OldCoder Actually, the *current* size is small. The SQLite3 is just 400MB. 20:56 OldCoder It is a *wide* world 20:56 OldCoder as opposed to physically large 20:56 OldCoder Wide and complex 20:56 OldCoder I will try doubling the minimum timeout *again* 20:57 OldCoder ACKed packet not in outgoing queue 20:57 OldCoder Opinions on that part vary ^ 20:57 OldCoder Most feel it is not the issue 20:58 OldCoder 20:58 OldCoder I don't think people can test, more than one person needs to understand this code. 20:58 OldCoder Otherwise, proller the troller is correct. 20:59 proller not always 20:59 OldCoder Hm? 20:59 OldCoder I got it to compile partway proller 20:59 OldCoder But there was a new Address class that messed things up 20:59 OldCoder Could not continue with the old code 20:59 Miner_48er proller what is your opinion? 20:59 OldCoder * can contest 21:00 celeron55 i personally hate sapier's version of the connection too 21:00 OldCoder I tried to revert 21:00 OldCoder Spent an hour on it 21:00 celeron55 but can't really do anything to it unless someone throws an unarguably better version at my face 21:00 OldCoder No lockups is better 21:01 Miner_48er what was the results of the revert? 21:01 OldCoder celeron55, you are more objective than proller... what is it that you dislike about the commits? 21:01 OldCoder Miner_48er, failure to compile 21:01 celeron55 it's too complex to be maintainable 21:01 OldCoder Agreed 21:01 celeron55 the issue is simple as that 21:02 OldCoder proller, Miner_48er I was able to handle the obvious missing symbol errors. But there was a class change that seemed fundamental. 21:02 OldCoder Going from u16 for "address" to an Address class that had both a 32 bit address and a port number 21:02 OldCoder The correct changes were not clear 21:03 celeron55 can't you just copy-paste the stuff from an old version instead of trying any kind of merge? 21:03 OldCoder I reverted Sapier's commit as proller suggested 21:03 OldCoder Then tried to patch the result up to compile 21:03 OldCoder The Address class seems to be needed 21:03 celeron55 yeah i mean don't use git 21:03 OldCoder Go on 21:04 OldCoder The interface between the connection code and the rest breaks regardless 21:04 celeron55 just take the code from an old version and manually replace stuff with it in HEAD 21:04 celeron55 does it break so much? 21:04 OldCoder Not for a knowledgeable core dev 21:04 celeron55 i would imagine it's pretty much intact 21:04 OldCoder I don't know what to do with this one Address class 21:05 OldCoder If anybody can comment on that I may try again 21:06 celeron55 i can't comment anything based on this information 21:06 OldCoder I figured, thanks 21:06 Miner_48er I wonder what would happen if you have saved the world with WE and pasted it in a new world since SC is you only world experiencing this 21:06 OldCoder Others are 21:06 celeron55 if you can point me to a commit or compiler errors, then maybe 21:06 Miner_48er *your 21:06 OldCoder And you yourself have seen reports elsewhere 21:06 OldCoder celeron55, I will research further and report back 21:06 OldCoder Address &address 21:06 OldCoder That class seems to be new 21:07 celeron55 the Address class has always existed 21:07 OldCoder There is breakage related to that. I will research further. 21:07 celeron55 it just has had some extra functionality added to it 21:07 OldCoder One question only, then 21:07 OldCoder If there is code that formerly used u16 for address and now uses Address, how do I reconcile that? 21:08 celeron55 i'm not aware of code that has used u16 for an address, that doesn't make sense to me 21:08 OldCoder I will withhold further questions for now, then. However, thank you. 21:08 celeron55 do you mean a peer id? 21:09 proller OldCoder, you can look to freeminer/next - it have fully replaced connection 21:09 OldCoder No, I don't think so. This happened after the revert and there were parameters for addresses. I will wait until there is a specific error message again to show. 21:09 OldCoder But it isn't compatible with MT, is it? 21:09 OldCoder If I got up to speed, I could port it 21:09 proller no, but you can look how anything ported 21:09 OldCoder But I'm already a week behind with work in real life. All right, that is all for now. 21:09 celeron55 well, proller has to publish it under LGPL first 21:10 OldCoder There is a license dispute, I take it 21:10 OldCoder But concepts are free 21:10 celeron55 yes of course 21:10 proller celeron55, look != commit to minetest 21:10 celeron55 there's just a much larger amount of work than concepts 21:10 OldCoder Indeed 21:11 OldCoder error: no matching function for call to âAddress::Address(irr::u16&) 21:11 OldCoder This is the sort of error I'm seeing 21:14 proller 85fe75d1cbaf372f0a98558adcd5a612a0e19602 21:14 OldCoder ? 21:14 OldCoder Is that different than the other one? 21:15 proller ups, miss 21:15 OldCoder proller, give me Sapier's commit again please 21:15 OldCoder I can't find it in scrollback 21:15 OldCoder Ah 21:15 OldCoder there it is 21:24 jin_xi hm, whats up with lines 412/413 in connection? its just whitespace but looks very wrong 21:24 jin_xi https://github.com/minetest/minetest/blob/master/src/connection.cpp#L412 21:31 kahrl jin_xi: probably because of crazy people who think a tab is 4 spaces :P 21:36 proller its wrong from line 40 to line 3126 21:41 OldCoder proller, yes, we get your feelings about that. I am trying again to revert. 22:51 OldCoder for argument 1 from âirr::u16 {aka short unsigned int}â to âAddress 22:51 OldCoder celeron55, this is what I was referring to ^ 22:51 OldCoder How do I convert an irr:u16 to an Address? 22:55 OldCoder Hm 22:55 OldCoder Pasting old code 23:24 OldCoder I am unable to revert Sapier's changes but found a possible gap in the code 23:24 OldCoder Trying a new patch